Senior Research

Political Instability and Thailand’s Economic Growth

Gayvaleen Kijkul

534 58118 29

Advisor: Somprawin Manprasert, Ph.D.

November 29, 2013

Senior Research Submitted in Partial Fulfillment of the Requirements for the

Bachelor of Arts degree in Economics (International Program)

The Bachelor of Arts Program in Economics

Faculty of Economics

Chulalongkorn University

Academic Year 2013

Approve

_______________________________________

(Assoc. Prof. Sothitorn Mallikamas, Ph.D.)

Chairman

Date of Approval _________________________

Abstract

This paper examines the relationship between political instability and economic

growth in Thailand from 1976 to 2010. Generally, there exists a negative endogenous

relationship between political instability and economic growth: political instability deters

economic growth, and a lack of economic growth could spark political instability.

I measured political instability by constructing a socio-political instability (SPI) index

using the principal components method. Estimating a system of simultaneous equations, I

found that the relationship between political instability and economic growth in Thailand

over the specified time period was insignificant. The main driver of the Thai economy (in the

past and in the present) is trade, or more specifically exports, and as long as political

instability does not disrupt the flow of exports, it will not have a significant effect on

economic growth. In the future, however, if Thailand were to move away from an export-

driven economy, the government will have to ensure that political instability has a minimal

effect on whichever factor the Thai economy becomes dependent on.

Table of Contents

Introduction............................................................................................................................................. 5

The Relationship between Instability and Growth .............................................................................. 5

Thailand: Political Turmoil and Economic Growth ........................................................................... 5

Research Objectives ................................................................................................................................ 7

Hypotheses .............................................................................................................................................. 7

Scope ....................................................................................................................................................... 7

The Literature Review ............................................................................................................................. 8

Introduction......................................................................................................................................... 8

Review Strategy ................................................................................................................................... 8

The Effect of Political Instability on Economic Growth: A Review of the Literature ......................... 8

Conclusion and Contribution ............................................................................................................ 12

Conceptual Framework ........................................................................................................................ 13

Methods and Procedures ...................................................................................................................... 14

Data .................................................................................................................................................. 14

The Socio-Political Instability Index ................................................................................................. 14

A System of Simultaneous Equations ................................................................................................ 15

Results ................................................................................................................................................... 17

Discussion of Results ............................................................................................................................ 18

Thailand: An Export Driven Country ............................................................................................... 18

Instability and Exports ...................................................................................................................... 18

Instability and Growth ...................................................................................................................... 19

The 2008 Suvarnabhumi Shutdown – An Event Study ...................................................................... 21

Conclusion ............................................................................................................................................ 23

Policy Implications ........................................................................................................................... 23

Limitations and Extensions ............................................................................................................... 23

Reference List........................................................................................................................................ 24

Appendix ............................................................................................................................................... 26

Appendix A – Socio-Political Instability (SPI) Index ........................................................................ 26

Appendix B – Regional Growth (RGROWTH) .................................................................................. 27

4

Table of Figures

Figure 1: Thailand's Economic Growth (1960-2012) ............................................................................. 5

Figure 2: Conceptual Framework ......................................................................................................... 13

Figure 3: Representation of Socio-Political Instability Index ............................................................... 14

Figure 4: An Index of Socio-Political Instability (1976-2010) ............................................................. 15

Figure 5: Representation of System of Simultaneous Equations .......................................................... 15

Figure 6: Estimation Results of Growth Equation ................................................................................ 17

Figure 7: Estimation Results of Instability Equation ............................................................................ 17

Figure 8: Thailand's Exports and Growth ............................................................................................. 18

Figure 9: Thailand's Exports and Political Instability ........................................................................... 19

Figure 10: Monthly Value of Thailand's Exports (2006-2010) ............................................................. 21

Figure 11: Regression Based on Reference Index ................................................................................ 22

Figure 12: Event Study of Suvarnabhumi Shutdown (Nov-Dec 2008) ................................................ 22

5

Introduction

The Relationship between Instability and Growth

The relationship between political instability and economic growth has been well

documented in the field of the political economy. There are several valid explanations to

explain the negative relationship between instability and growth. Political instability results in

uncertainty regarding a country’s future. Oftentimes, property rights are threatened and laws

are overthrown. This hinders productive activities such as investment and consumption as

economic agents are generally risk averse, and ultimately lowers economic growth. The

majority of academic papers in the field utilize panel data of a large number of countries over

a certain time period to examine whether or not political instability deters growth. Many note

the possibility of reverse causation between the two variables and account from problems of

endogeneity in their studies, since it is very likely that instability deters growth, but low

growth could instill instability as well.

Thailand: Political Turmoil and Economic Growth

To date, the Kingdom of Thailand has had a total of seventeen charters and

constitutions. From the very first official constitution established on December 10th

, 1932 –

the date that is celebrated annually as Thailand’s Constitution Day for moving the country

from an absolute to a constitutional monarchy – to the current 2007 Constitution of Thailand,

Thailand claims the title of being one of the countries with the most constitutional changes in

the world. Furthermore, the country boasts a total of eighteen military coups d’état since 1932,

a statistic that once again ranks it among the countries which have had the most coups in the

world.

Figure 1: Thailand's Economic Growth (1960-2012)

Historically, Thailand has enjoyed healthy rates of annual economic growth that

typically range from 5 to 10 percent, as shown in Figure 1 above. Gross domestic product in

both absolute and growth terms has grown continuously over the past half century, except for

-15

-10

-5

0

5

10

15

0

5E+10

1E+11

1.5E+11

2E+11

2.5E+11

3E+11

3.5E+11

4E+11

1960 1970 1980 1990 2000 2010

GD

P G

row

th (

%)

GD

P (

Cu

rre

nt

US$

)

GDPGROWTH 1997 Asian

Financial Crisis

2008 Subprime

Mortgage Crisis

6

two incidences where the Thai economy experienced negative growth: the 1997 Asian

Financial Crisis and the 2008 Subprime Mortgage Crisis. These two events have been

highlighted in yellow in Figure 1. Otherwise, however, it would be fair to say that Thailand’s

economy has continued to grow over the past years despite an above-average frequency of

government changes and military coups.

It is hard to deny that Thailand is a politically unstable country, but yet its economy

has continued to grow and thrive. Is the Thai economy not growing at its full potential? What

are the causes of these constant governmental changes and military coups, and what effects

do they entail? Does political instability deter economic growth, or do poor economic

performances encourage political instability? Or does the relationship exist both ways? In this

paper, I aim to examine the relationship between Thailand’s political instability, quantified by

an index of socio-political instability (SPI), and economic growth, as measured in growth in

gross domestic product.

7

Research Objectives

The main objective of this paper is to study the relationship between Thailand’s

political instability and economic growth. More specifically, what effect does political

instability – defined by an SPI index – have on economic growth as measured in growth of

gross domestic product, and vice versa. In the particular case of Thailand, are these two

variables jointly interactive and endogenous? Or does the effect only exist in one direction?

Furthermore, is political stability an important driver of Thai economic growth? If not, what

are the main drivers of Thailand’s economy? I hope to answer these questions in my study.

Hypotheses

As has been proven in previous studies, I expect a negative relationship between

political instability and economic growth. I hypothesize that a high level of political

instability, as indicated by a high value in the SPI index, will hinder economic growth.

However, for the case of Thailand, I do not think that poor economic performances

significantly increase political instability. This is because I see that government changes and

military coups have been frequents occurrence despite the general upward trend of the

country’s economic growth. In other words, I expect a significant negative relationship

between the two variables, but only in the direction from instability to growth, and not vice

versa.

Scope

Given the constraints on time and resources, I have chosen to narrow down this study

to focus solely on the Kingdom of Thailand. I have chosen Thailand not only because it is my

country of birth and residence, but also for its infamous history of military coups and

government changes. I do not study all possible forms of political instability, but rather just

political instability as quantified in an index of socio-political instability consisting of three

variables: the number of government changes, the number of military coup d’états, and major

episodes of political violence in a given year. As for the time period, my study covers the

years 1976-2010, which is the largest continuous range possible for which there is available

data for all the variables in the study.

8

The Literature Review

Introduction

In this literature review, I explore the vast literature concerning the relationship

between political instability and economic growth. The structure of this literature review is as

follows. To start off, I explain the review strategy I used to narrow down my choice of

academic papers. The review itself is topic-based. First, I briefly discuss the various models

that have attempted to explain economic growth and some determinants that have been found

to stimulate growth. Second, I discuss the results and methods used by numerous papers to

study the relationship between instability and growth. The last section is a conclusion where I

also put forth my proposed contribution.

Review Strategy

In the process of selecting which papers to include in this literature review, I used

keywords such as “political instability and economic growth”, “determinants of economic

growth”, and “political instability in Thailand” in the Google Scholar search engine, which

ultimately led me to formal online sources like Scopus, ScienceDirect, and JSTOR. From the

wide array of available papers, I narrowed my choice down to a number of academic papers

using several methods. I chose papers with high citation counts because they were an

indication of classic works in the field. By reading the provided abstract of each paper, I

decided whether or not they were relevant to my research question. Sometimes, reading the

whole introduction of a paper rather than just its abstract was necessary for me to be able to

judge its relevance. I was particularly selective in terms of the publication date for papers

regarding political instability in Thailand; I preferred recently published papers in order to

account for the most recent political instability in the past years. Starting from a small

selection of papers, I referred to the core citations of each paper and broadened my choice of

literature accordingly. Furthermore, in online journal databases such as Scopus, related

documents and references are often listed alongside papers. Another very useful tool that

helped me in my review strategy was the JEL (Journal of Economic Literature) classification

code system for academic papers in the field of economics. The majority of the papers I

reviewed were coded ‘O’ for ‘Economic Development, Technological Change, and Growth’,

specifically ‘O1’ for ‘Economic Development’ and ‘O4’ for ‘Economic Growth and

Aggregate Productivity’. ‘D7’ for ‘Analysis of Collective Decision-Making’ under ‘D’ for

‘Microeconomics’ and ‘P16’ for ‘Political Economy’ were also recurring tags in the literature.

Lastly, given the limitation of resource availability, I could not access all the papers I wanted

to because Chulalongkorn was not subscribed to all journals. The papers I reviewed were the

ones accessible via Chulalongkorn’s current subscriptions.

The Effect of Political Instability on Economic Growth: A Review of the Literature

There is a vast amount of literature concerning economic growth. From the

neoclassical model of growth to the endogenous models of growth, different economists have

put forth different arguments in attempts to identify and explain the factors that push an

economy to grow. For example, in Barro’s (1996) cross-country empirical study of around

100 countries from 1960 to 1990, he finds several determinants of economic growth. He finds

that economic growth is positively affected by higher life expectancy, higher initial schooling,

better maintenance of the rule of law, improvements in the terms of trade, lower fertility,

9

lower inflation, and lower government consumption. An increase in political freedom has

only a weak effect on growth and is probably nonlinear; an initial increase in political rights

increases growth, but once a certain level of democracy has been reached, further increases in

political freedom might actually decrease growth. Also, for given values of these variables,

there is an inverse relationship between growth and the initial level of GDP per capita. This

supports the notion of conditional convergence put forth by the neoclassical model of growth:

the lower the initial level of real GDP per capita, the higher the expected growth rate and vice

versa due to diminishing returns to capital.

However, economic variables are not the sole determinants of a country’s growth.

Political-economy literature suggests that differences in growth and economic outcomes

across countries are better explained through both economics and politics. Shleifer and

Vishny (1993) and Mauro (1995) find that corruption is detrimental to economic growth.

Murphy, Shleifer, and Vishny (1991) find that rent-seeking activities, especially in weak

governments vulnerable to lobbying and pressure groups due to the threat of losing office,

lower economic growth. Acemoglu et al. (2003) argue that distortionary and unsustainable

macroeconomic policies are not the main causes of economic volatility and crises but rather

symptoms of deeper institutional problems.

More specifically, a broad range of studies have examined the relationship between

political instability and economic growth. Most find that political instability has a negative

effect on economic growth for many plausible reasons. Political instability is often associated

with violence and illegal acts such as riots, political assassination, and military coups. This

poses a direct threat to property rights. It also results in an economic environment of

uncertainty which might induce economic agents to lower savings and investment which in

turn lower growth. Furthermore, the possibility of soon being replaced in office for

policymakers might result in short-sighted macroeconomic policies. More direct effects of

political instability on economic growth include the emigration of valuable human capital and

the destruction of infrastructure.

Different studies adopt different approaches of measuring political instability. Mainly,

there are two commonly used methods. The first method focuses on executive instability, or

the frequency of government changes and collapses. Alesina et al. (1996) define political

instability as the propensity of change in the executive power. The variable GCHANGE is 1

in years with both regular and irregular transfers of the executive power and 0 in years with

no government change. The variable MJCHANGE captures all irregular and regular

government changes that imply substantial changes in the political ideology of the

government. The variable COUP captures the occurrence of only irregular transfers of power,

such as coups d’etat. Chen and Feng (1996) defined political instability as the probability of

regime change, a continuous variable found using the limited dependent variable estimation

method. Feng (1997) stressed the need to differentiate between regime stability and

government stability, pointing out that not doing so could lead to ambiguous and

contradicting results as have been found by certain past studies. He differentiated between

irregular political changes, major regular political changes, and minor regular political

changes. Cukierman et al. (1992) and Edwards and Tabellini (1991) also use this definition of

political instability as the propensity to observe government changes in their studies of the

effects of political instability on inflation.

The second method commonly used to define political instability is the use of socio-

political indices. Starting with a list of variables related to political violence and social unrest,

an aggregate index is constructed. One statistical technique that can be used to reduce a

10

multidimensional set of variables to just one variable in one dimension is the method of

principle components. Alesina and Perotti (1996) employ a socio-political instability index

which aims to capture the occurrence of violent incidents of political unrest. The index is

constructed using the method of principle components on variables such as the number of

politically motivated assassinations, the number of people killed in phenomena of domestic

mass violence, the number of successful, and the number of attempted but unsuccessful coups.

Barro (1991) uses the number of revolutions and coups per year to build an index of political

instability. Ades and Chua (1997) builds a regional instability index using the average

number of revolutions and coups per year of a country’s neighboring countries. Hibbs (1973),

Gupta (1990), Barro (1991), Ozler and Tabellini (1991), and Benhabib and Spiegl (1992) use

similar indices in their studies on the effect of political instability on growth, savings, or

investment.

Alesina and Perotti (1996) point out the advantages and disadvantages of both

approaches. For any given level of expected government turnover, social unrest does not have

any direct effect on political uncertainty. This implies that the socio-political instability

indexes used in many studies are inaccurate ways of measuring political instability. On the

contrary, advocates of socio-political instability indexes reason that social unrest affects

economic growth for reasons other than just the high expectation of a change in the executive

power. Disorder and unrest have direct effects on productivity and investment: physical

safety is threatened and politicians might increase taxes on capital due to their shortened

horizon. Yet, these are not the only two methods in defining political instability. Jong-A-Pin

(2009) uses exploratory factor analysis rather than principle components analysis and

identifies four distinguishable dimensions of political instability: politically motivated

violence, mass political violence, instability within the political regime, and instability of the

political regime.

Numerous studies find evidence to support to negative relationship between political

instability and economic growth. Alesina et al. (1996) find that political instability

significantly lowers growth. Unconstitutional executive changes such as coups d’etat have the

strongest adverse effect on economic growth, whereas regular turnovers of the government

have less strong effects. This applies to both democratic and non-democratic countries.

Alesina and Perotti’s (1996) study reveals that income inequality increases socio-political

instability which then reduces investment and economic growth. When income inequality is

high, the large group of worse off individuals has a higher possibility of feeling frustrated at

the small group of better off individuals and the current socio-economic situation, and is thus

more likely to demand radical changes that might result in mass violence and illegal seizures

of power. A rise in political instability decreases investment through three channels. First, the

expected level of taxation of factors that can be accumulated increases. Second, productive

activities are disrupted, causing the productivity of labor and capital to fall. Third, uncertainty

increases, inducing economic agents to postpone investment, invest abroad, or consume more

instead. The fall in investment lowers growth. Chen and Feng (1996) find that political

assassinations and the possibility of regime change have a negative effect on economic

growth. Feng (1997) concludes that the possibility of regime change, political assassinations,

and lack of economic freedom all lead to a decrease in economic growth. Jong-A-Pin (2009)

finds that the instability of a political regime reduces growth because investors feel insecure

and uncertain about property rights. Instability within a political regime, however, implies a

relatively high level of political competition and increases economic growth because

incompetent politicians are likely to be replaced by more competent ones. Ades and Chua

(1997) put forth the idea that political instability spills over across geographical areas.

Regional instability measured by political instability in neighboring countries affects growth

11

negatively, and interestingly has as much of an effect as domestic political instability. Barro

(1991) finds that political instability is negatively correlated to growth as well.

However, not all studies find that political instability and economic growth have a

negative relationship. Olson (1982) finds that governments that remain in office for longer

periods of time are more likely to engage in suboptimal policies because they have the

tendency to please certain interest groups. Huntington (1968) suggests a nonlinear

relationship between instability and growth, depending on the country’s level of development.

Instability and growth have a negative relationship in richer countries that have institutions

ready to deal with the social and economic changes brought about by growth. However, the

relationship might be positive in poorer countries; when poor economies experience rapid

growth, new demands are generated and social unrest may actually increase. Londregan and

Poole (1990) find that neither the current propensity for a coup or a past record of coups

significantly affects a country’s growth rate, suggesting that political instability does not

reduce growth.

One important issue to keep in mind is the possibility that economic growth and

political instability are jointly endogenous. An increase in political instability might cause a

fall in economic growth, but a fall in economic growth might induce increased political

instability as well. Alesina et al. (1996) tackle this problem of endogeneity by estimating a

system of two equations. The first equation is a probit regression that estimates the propensity

of a change in the government, and the second equation is a regression for economic growth.

They find that interaction between the two variables could potentially lead to a vicious circle

where a country is socio-politically unstable because it isn’t growing, and isn’t growing

because it is socio-politically unstable. Unsatisfactory rates of economic growth instill a

sense of dissatisfaction in the population which might lead to political unrest. However, they

find that the low economic growth only significantly affects government collapses in the

form of coups d’etat but not otherwise. Jong-A-Pin (2009) similarly finds that the two

variables exhibit reverse causality and feedback effects: the instability of a political regime

reduces economic growth, but a lack of economic growth creates instability within the

political regime as well. Additionally, many studies have proven that growth has an effect on

political instability. Londregan and Poole (1990) suggest that satisfactory rates of growth

significantly reduce the chances of a coup and low rates of growth increase political

instability. Feng (1997) finds that economic growth reduces the occurrence of coups and

major irregular changes of the government and increases the occurrence of minor regular

changes. Intuitively, low growth increases the chances of an upset population overthrowing

the government, whereas high growth results in small reshufflings in a government rather

than major changes because past results have been satisfactory, meaning the new regime is

similar and very much aligned to the old regime with very minor changes.

There are many possible channels linking political instability and economic growth.

Alesina and Perotti (1996) show that investment is the channel linking instability and growth.

Ades and Chua (1997) find that there are two main channels through which regional

instability lowers growth. First, a disruption of trade flows might result from a blockage of

trading routes and the destruction of transportation networks. Second, domestic military

outlays might have to be increased in order to maintain border security and prevent the

instability from spreading, thus crowding out investment and more productive activities that

enhance economic growth. Aisen and Viega (2010) find that instability affects growth mainly

via a decrease in total factor productivity growth, and also via falls in physical and human

accumulation to a smaller effect.

12

Many studies regarding the relationship between political instability and economic

growth take into consideration the effect of expectations. For example, Londregan and Poole

(1990), Alesina et al. (1996), Chen and Feng (1996), and Feng (1997) all use the propensity

for or possibility of political instability as variables in their models rather than just actual

occurrences. The effects of political instability on growth are ex-ante rather than ex-post;

even if there is ultimately no actual change in the political regime, mere expectations of

regime instability are enough to influence growth.

Finally, the results of various studies indicate that political instability is fairly

persistent. Alesina et al. (1996) find that the past occurrence of government changes increases

the chances of government changes in the future. Along the same lines, Londregan and Poole

(1990) find that coups d’etat have substantial political aftereffects. A successful coup

increases the likelihood of another coup for up to six years. When a government is

overthrown, it becomes much more likely for a coup to occur during the subsequent

government’s term.

Conclusion and Contribution

In conclusion, there is substantial evidence that there exists a negative relationship

between political instability and a country’s economic growth. It is very likely that these two

variables exhibit reverse causality, and so the problem of endogeneity has to be carefully

considered when constructing a model. It has also been found that the mere expectation of

political instability affects growth rates, and that political instability is relatively persistent.

Despite the inexhaustible list of works concerning the relationship political instability

and economic growth, I hope to contribute by specifically studying this relationship in the

context of Thailand’s sociopolitical environment. Most studies employ panel data of a huge

number of countries over a certain time period, but I solely focus on Thailand in an attempt to

shed some light on the country’s intriguing problem of continuous political instability.

13

Conceptual Framework

The determinants of economic growth are numerous: free trade, capital accumulation,

saving, technology, and productivity just to name a few. However, this paper focuses on the

relationship between political instability and economic growth. In this study, political

instability is gauged by an socio-political instability index.

Political instability negatively affects growth by increasing uncertainty in the

economic environment. A high probability of a change in the government leaves the future of

property rights, laws, and policies unclear. This could have an adverse impact on productive

economic decisions: people will most likely save and invest less, or might even prefer to

invest abroad instead. As news of political instability spread worldwide, foreign investors

will most likely avoid putting their money in countries they deem politically unstable.

Conversely, a lack of economic growth could potentially lead to an increase in

political instability. There are two relevant explanations. Firstly, in countries where growth

increases income inequality, growth could instill a sense of anger and dissatisfaction among

the large group of people with relatively low income, making a coup d’état more likely.

Secondly, extremely poor growth rates could result in the general public being unsatisfied

with the current government, increasing incentives for antigovernment activities, and making

political instability more likely.

Since there are valid arguments for both directions of the relationship, the interaction

between the two variables could ultimately lead to a vicious cycle. Instability lowers growth,

but low growth increases instability as well.

POLITICAL INSTABILITY

SPI Index

ECONOMIC GROWTH

% Change in GDP

Investment and FDI

Savings

Productivity

Capital Accumulation

Trade Flows

Increased Military Outlays

Income Inequality

General Dissatisfaction

Investment

Education

Trade

Global and

regional growth

Figure 2: Conceptual Framework

14

Methods and Procedures

Data

All the data for the variables used to estimate the system of simultaneous equations

was obtained from the World Bank Open Database (World Bank, 2013). Data used to

construct the socio-political instability (SPI) index was obtained from both the website of the

Thai Secretariat of the Cabinet and the data page of the website for the Integrated Network

for Societal Conflict Research (Marshall, 2013).

The Socio-Political Instability Index

The socio-political instability index used in this study was constructed using the

method of principle components, or principal component analysis. Given a set of variables

that are possibly correlated, this computational procedure yields sets of values that represent

different dimensions of the combined variables. I chose to include three variables in my index:

the number of changes in Prime Minister per year (GCHANGE), the number of coups d’état

per year (COUP), and a measure of major episodes of political violence (ACTOTAL). It is

useful to note that the last variable, ACTOTAL, is already an index in itself, constructed by

the Center for Systemic Peace and Integrated Network for Societal Conflict Research to

reflect political violence across different countries. For more information regarding the

variables used in my index, please refer to Appendix A.

I chose to include these three variables in my index for a number of reasons. First, the

number of government changes (GCHANGE) – in this case, specifically the change of prime

ministers – and the number of coups per year (COUP) are common measure of political

instability. A frequent change of prime ministers instills uncertainty in the economy as

economic agents are not sure whether or not policies will change. Coups are perceived as

threats to legal and property rights, thus discouraging productive economic activities. Second,

ACTOTAL is an index in itself that covers various incidences of political violence, such as

civil violence, civil violence, and ethnic violence.

After inputting the data for the three variables into Gretl (econometrics software) and

using the principal component analysis function, I was able to obtain the following index for

socio-political instability.

SPI = 0.700(GCHANGE) + 0.701(COUP) - 0.133(ACTOTAL)

Figure 3: Representation of Socio-Political Instability Index

15

Figure 4: An Index of Socio-Political Instability (1976-2010)

Figure 3 is a graphical representation of the obtained SPI index from 1976 to 2010,

the time period of this study. Particularly high values for the SPI index (2.8 for the year 1992

and 2.7 for the year 2008) can be justified by the fact that three different prime ministers

holding office in each of these two years. Years in which there were coups, such as 1991 and 2006,

also have relatively high SPI values.

A System of Simultaneous Equations

I decided to use a system of two simultaneous equations to account for the possible

endogeneity and feedback effects between political instability and economic growth. The two

equations are inputted into Gretl and the coefficients are estimated based on the Full

Information Maximum Likelihood (FIML) method. The basic form of my proposed

econometric model is shown below.

GROWTH = α + β1 (SPI)

+ β2 (GROWTH-1) + β3 (RGROWTH) + β4 (TRADE) + β5 (INVEST)

+ β6 (EDUC)

SPI = α + β1 (GROWTH)

+ β2 (GROWTH-1) + β3 (RGROWTH) + β4 (TRADE) + β5 (INVEST)

+ β6 (GCHANGE-1)

In the growth equation, the main relationship of interest is between growth

(GROWTH), the dependent variable, and political instability (SPI), an independent variable.

Other explanatory variables in the equation are lagged growth (GROWTH-1), regional growth

(RGROWTH, a weighted average of the GDP growths of ASEAN countries – refer to

Appendix B for a detailed explanation), the percentage change in the total value of imports

-1

-0.5

0

0.5

1

1.5

2

2.5

3

1970 1975 1980 1985 1990 1995 2000 2005 2010 2015

Soci

o-P

olit

ical

In

stab

ility

(SP

I) I

nd

ex

Figure 5: Representation of System of Simultaneous Equations

16

and exports in current US dollars (TRADE), and the percentage change in the net value of

inflows of foreign direct investment in current US dollars (INVEST). Because the model is a

system of simultaneous equations, an instrumental variable is needed in each equation to

fulfill the conditions necessary for the estimation of the coefficients to be possible. An

instrumental variable has to affect one dependent variable, but not the other. For example, for

the growth equation, the instrumental variable has to directly affect economic growth

(GROWTH) but not political instability (SPI). In this case, I chose the enrollment rates in

primary schools (EDUC) as the instrumental variable for the growth equation because

education is known to directly affect economic growth (higher enrollment rates correspond to

higher rates of growth), but does not directly affect political instability.

The instability equation is constructed similarly. The main relationship of interest is

between the dependent variable of political instability (SPI) and the independent variable of

economic growth (GROWTH). Other explanatory variables are the same as those of the

growth equation, with the only exception being the instrumental variable. For the instability

equation, I used a lagged variable of government change (GCHANGE-1) as the instrumental

variable because a change in the prime minister in the previous year has an effect on this

year’s measure of political instability but does not have a direct effect on this year’s growth.

17

Results

The estimation results of the regressions are shown below.

coefficient std. error z p-value

const -4.1486 17.9645 -0.2309 0.8174

SPI -0.8974 6.2358 -0.1439 0.8856

GROWTH-1 0.1061 0.2316 0.4582 0.6468

RGROWTH 0.8596 0.1723 4.988 0.0000006

TRADE 8.4482 3.6943 2.287 0.0222

INVEST -0.4883 1.4420 -0.3387 0.7349

EDUC 0.0409 0.1791 0.2285 0.8193

Figure 6: Estimation Results of Growth Equation

coefficient std. error z p-value

const -0.0069 0.4372 -0.0158 0.9874

GROWTH -0.3472 0.7170 -0.4842 0.6283

GROWTH-1 0.0592 0.0763 0.7749 0.4384

RGROWTH 0.3002 0.6233 0.4817 0.6300

TRADE 2.9770 6.0881 0.4890 0.6249

INVEST -0.0233 0.5396 -0.0432 0.9656

GCHANGE-1 0.0907 0.4892 0.1854 0.8529

Figure 7: Estimation Results of Instability Equation

Surprisingly, the only two variables that were statistically significant were regional

growth (RGROWTH, at a 99% confidence interval) and trade (TRADE, at a 95% confidence

interval), both from the growth equation. As expected, an increase in regional growth and/or

the value of trade would cause an increase in domestic economic growth. More specifically,

the results imply that a 1% increase in regional growth would correspond with a 0.86%

growth in the Thai economy, assuming all else constant. Interpreting the coefficient of

TRADE in the growth equation, a 100% growth in exports would result in an 8.45% growth

in the Thai economy. In the political instability equation, none of the explanatory variables

were significant, resulting in a very low adjusted r-squared value.

***

**

Dependent Variable: GROWTH

Adjusted R-Squared: 0.7756

Dependent Variable: SPI

Adjusted R-Squared: -0.17349

18

Discussion of Results

Thailand: An Export Driven Country

The main driver of Thailand’s economic growth has always been exports. Historically,

exports have amounted to over two-thirds of Thailand’s GDP for several decades. For

example, in 2012, exports totaled 75% of Thailand’s GDP. As shown in the figure below,

fluctuations in exports (the red line) strongly correspond to fluctuations in economic growth

(the blue line).

Figure 8: Thailand's Exports and Growth

We have now established that there is a strong relationship between exports and

economic growth in Thailand. Intuitively, political instability would have a significant effect

on economic growth if it had a significant effect on Thailand’s exports. Thus, this is the

relationship we will examine next.

Instability and Exports

In the figure below, the values for exports and instability have been plotted over the

time period 1976 to 2010, which is the time scope of this study. Graphically, we can see that

there seems to be no obvious or clear relationship between the two variables. From 1976 to

around 1990, it seems as if both lines are moving in the same direction. However, it does not

make intuitive sense because this would imply that an increase in political instability

corresponds with higher exports. In the second half of the graph, the blue and red lines appear

to be extremely unrelated and almost random. Thus, it would be fair to say that political

instability and exports, in the case of Thailand from 1976 to 2010, are unrelated.

-30.00%

-20.00%

-10.00%

0.00%

10.00%

20.00%

30.00%

40.00%

50.00%

-15

-10

-5

0

5

10

15

1970 1975 1980 1985 1990 1995 2000 2005 2010 2015

Ch

ange

in T

rad

e/E

xpo

rts

(%)

Re

al G

DP

Gro

wth

(%

)

GROWTH

EXPORTCHANGE

TRADECHANGE

19

Figure 9: Thailand's Exports and Political Instability

Instability and Growth

In order for Thailand’s political instability to affect its economic growth, it has to

affect Thailand’s single main driver of economic growth: the exports of goods and services.

The regression results have indicated that trade (TRADE) is the major variable that

determines economic growth. The other statistically significant variable, regional growth

(RGROWTH), can be interpreted as other factors that affect growth but are not covered by

the other variables. For example, RGROWTH would capture factors such as the Asian

Financial Crisis of 1997 or the Subprime Mortgage Crisis of 2008. These events, without a

doubt, had a huge effect on Thailand’s economic growth, but cannot be captured by the other

explanatory variables.

A different explanation is needed for why a lack of economic growth does not spur

political instability. This sort of relationship has been documented in several studies as

referred to in the literature review, but does not seem to exist in Thailand. Looking at the 18

coups d’état, 4 of which are covered in the scope of this study, none of them had an obvious

economic justification. The coups weren’t results of disappointing economic growth, but

rather of political reasons and power struggles.

Furthermore, I hypothesize that the frequency and nature of military coups in the past

have altered expectations regarding the effects and consequences. Coups occur extremely

often, compared to other countries with developing economies similar to that of Thailand,

and do not have significant effects on the economy. The limited effects of coups in the past

have caused most economic agents to realize that a coup is not a warning of lowered growth

in the near future and do not alter their economic activities very much. There may be short

-1

-0.5

0

0.5

1

1.5

2

2.5

3

-20.00%

-10.00%

0.00%

10.00%

20.00%

30.00%

40.00%

50.00%

1975 1980 1985 1990 1995 2000 2005 2010

SPI

Ind

ex

Ch

ange

in E

xpo

rts

(%)

EXPORTCHANGEPSI2.1

20

term volatility in certain variables, such as investment and the Baht currency, due to a certain

degree of uncertainty, but they soon regain stability and return to their previous situations.

The important thing is that exports are what drive Thailand’s economy, and as long as

they are not affected substantially, political instability will not have a significant effect on

economic growth. This has been the case for Thailand in the past and the present, or at the

very least from 1976 to 2010, the time scope of this study. However, if certain conditions

change, political instability might have a negative effect on economic growth. I have picked

out two possible conditions where this might be the case: 1) if political instability were to

have a significant effect on exports and 2) if Thailand shifted away from being an export-

driven economy.

The first condition is if political instability were to have a significant effect on exports,

then it would definitely have a significant effect on the economy. For example, if political

demonstrations were to disrupt manufacturing, which accounts for up to 90% of exports,

exports would fall, ultimately causing economic growth to fall. This could be the case if

protesters closed down important manufacturing hubs, rather than major streets near

governmental offices in Bangkok or important business districts that are home to several

multinational corporations and offices but not factories. Another possible scenario would be

if political instability were to disrupt major transportation hubs such as airports or sea ports

responsible for the shipping of cargo to foreign countries. This would disturb the flow of

exports not only in the short term, but possibly the long term as well if trading partners suffer

from Thailand’s sudden inability to deliver its promised goods and lose trust or confidence.

An example of how political instability might affect an airport, and ultimately exports and

economic growth, is given in the next section.

The second condition is the case if Thailand shifted away from being an export-driven

economy. If the main driver of Thailand’s economy changed from exports to another factor

such as services, domestic consumption, or investment, it is possible that political instability

would deter economic growth. For example, if services became the main driver of the

economy and political demonstrations were to interfere with major commercial districts such

as Siam, Silom, or Sukhumvit, economic growth would be compromised. In other words,

although it might seem obvious, political instability will deter economic growth if it has an

effect on whatever factor the Thai economy is most dependent on.

21

The 2008 Suvarnabhumi Shutdown – An Event Study

Let us consider the shutdown of the Suvarnabhumi International Airport by the

People’s Alliance for Democracy (PAD), commonly known as the “Yellow Shirts”, from

November 25, 2008 to December 2, 2008. The Survarnabhumi airport was seized, closed

down, and no flights were allowed to arrive or depart. This is a clear example of condition 1 –

how political instability could significantly affect the economy by disrupting the flow of

exports.

Figure 10: Monthly Value of Thailand's Exports (2006-2010)

Judging by the graph above, it seems like exports suffered a huge blow in the months

of November and December in 2008 (last two data points of the red line) when compared to

the export values of the same months in other years. However, more analysis is needed to see

whether or not this drop is statistically significant. Daily data on exports would be ideal,

considering the 8-day time span of the airport shutdown. However, due to data unavailability,

I conducted this event study based on monthly data available from the website of the Bank of

Thailand.

I decided to use the data from the four years preceding 2008 (2005, 2006, 2007, and

2008) and the four years after 2008 (2009, 2010, 2011, and 2012) as a reference for the

expected rates of monthly export growth. I calculated the monthly export growth rates for all

the reference years and averaged them to construct the reference index. Lining up the growth

rates from January to October of the reference index to the actual growth rates in 2008, I

estimated a simple regression to capture the relationship between the reference index and the

actual growth rates in 2008. The resulting regression is as follows:

300,000.00

350,000.00

400,000.00

450,000.00

500,000.00

550,000.00

600,000.00

Jan Feb Mar Apr May Jun Jul Aug Sept Oct Nov Dec

Exp

ort

s (M

illio

n B

aht)

2006

2007

2008

2009

2010

22

EXPORT2008 = -0.005390 + 0.6183(EXPORTref)

Reference Actual (2008)

Jan - -

Feb 7.85% -8.75%

Mar 11.60% 11.48%

Apr -14.47% -9.76%

May 14.76% 13.23%

Jun 3.47% 6.93%

Jul -1.35% 8.44%

Aug 5.09% -6.22%

Sept 3.55% 1.95%

Oct -4.53% -6.09%

Figure 11: Regression Based on Reference Index

Using this regression, I calculated the expected monthly growth rates in exports for

the remaining two months of the year – November and December – for 2008, the two months

of interest because this is when the airport was seized. Then, I calculated the error which is

the difference between the expected value and the actual value that prevailed in 2008. Export

growth in November 2008 was 20.54% below the expected growth rate. To test for this

value’s statistical significance, I calculated the t-values by dividing the error term by the

standard error of the regression (calculated to be around 0.078). Results show that the fall in

export growth in November 2008 is statistically significant at a 97% confidence interval.

Reference Expected (2008) Actual (2008) Error t-test

Nov -1.56% -1.50% -22.04% -20.54% -2.6324

Dec -3.24% -2.54% 0.41% 2.95% 0.3780

Figure 12: Event Study of Suvarnabhumi Shutdown (Nov-Dec 2008)

This event study proves that political instability has a significant effect on economic

growth in the case where the flow of exports is disrupted. The People’s Alliance for

Democracy’s (PAD) seizure of the Suvarnabhumi Airport prohibited incoming and outgoing

flights to and from Thailand for both passengers and freight. As a result, export growth in

November 2008 was 20.54% below its expected level, and as long as Thailand’s economy is

heavily dependent on exports, a sizeable fall in exports will definitely compromise economic

growth.

23

Conclusion

In conclusion, there exists no significant relationship between political instability (as

defined by my socio-political instability index) and economic growth in Thailand in the past

and present, or at the very least from 1976 to 2010. Trade and regional growth have the most

significant effects on domestic economic growth; the Thai economy is most affected by the

volume of exports in a given year and external factors such as regional or global economic

crises. As long as political instability does not affect exports, the main driver of the Thai

economy, it will not affect growth. Intuitively, political instability will have an effect on

economic growth if 1) instability disrupts exports under the current export-driven economy

(as demonstrated in the event study of the 2008 Suvarnabhumi shutdown) or if 2) Thailand

shifts away from an export-driven economy and instability affects whichever factor the

economy becomes dependent on.

Policy Implications

These results have several policy implications for the present and the future. For the

present, the government should ensure that political instability does not disrupt trade and the

flow of exports by making sure that manufacturers and transportation hubs can operate as

normally as possible during times of political turbulence. As for the future, the government

will have to tailor its policies depending on how the Thai economy evolves; whatever the

economy becomes dependent on (i.e. services, investment, tourism, etc.), the government has

to make sure that political instability will have a minimal effect on those factors.

Limitations and Extensions

In this study, I employed annual data for all variables. I think it would be interesting

to do the same analysis using daily or monthly data instead to see whether or not the

relationship between political instability and economic growth might have a more short-term

or temporary significance. However, the availability of short-term macroeconomic variables

is a potential concern; these variables are usually reported on a yearly or quarterly basis.

Furthermore, it would be useful to explore alternative definitions to “political

instability”. I defined it in a very specific manner in this paper, constructing a socio-political

instability index, and defining it in different ways might provide additional insights.

24

Reference List

Acemoglu, D., Johnson, S., Robinson, J., & Thaicharoen, Y. (2003). Institutional causes,

macroeconomic symptoms: volatility, crises and growth. Journal of monetary

economics, 50(1), 49-123.

Ades, A., & Chua, H. B. (1997). Thy neighbor's curse: regional instability and economic

growth. Journal of Economic Growth, 2(3), 279-304.

Aisen, A., & Veiga, F. (2010). How does political instability affect economic growth?.

Alesina, A., Özler, S., Roubini, N., & Swagel, P. (1996). Political instability and economic

growth. Journal of Economic growth, 1(2), 189-211.

Alesina, A., & Perotti, R. (1996). Income distribution, political instability, and

investment. European Economic Review, 40(6), 1203-1228.

Alesina, A., & Perotti, R. (1994). The political economy of growth: a critical survey of the

recent literature. The World Bank Economic Review, 8(3), 351-371.

Bank of Thailand. (2013). International Trade. Available:

http://www.bot.or.th/ENGLISH/STATISTICS/ECONOMICANDFINANCIAL/EXTE

RNALSECTOR/Pages/StatInternationalTrade.aspx. Last accessed 12th Dec 2013.

Barro, R. J. (1996). Democracy and growth. Journal of economic growth, 1(1), 1-27.

Barro, R. J. (1991). Economic growth in a cross section of countries. The quarterly journal of

economics, 106(2), 407-443.

Benhabib, J., & Spiegel, M. (1992). The role of human capital and political instability in

economic development. CV Starr Center for Applied Economics, New York

University, Faculty of Arts and Science, Department of Economics.

Chen, B., & Feng, Y. (1996). Some political determinants of economic growth: Theory and

empirical implications. European Journal of Political Economy,12(4), 609-627.

Cukierman, A., Edwards, S., & Tabellini, G. (1989). Seigniorage and political instability (No.

w3199). National Bureau of Economic Research.

Edwards, S., & Tabellini, G. (1991). Political instability, political weakness and inflation: an

empirical analysis (No. w3721). National Bureau of Economic Research.

Feng, Y. (1997). Democracy, political stability and economic growth. British Journal of

Political Science, 27(03), 391-418.

Gupta, D. K. (1990). The economics of political violence. New York: Praeger.

25

Hibbs, D. A. (1973). Mass political violence: A cross-national causal analysis(Vol. 253).

New York: Wiley.

Huntington, S. (1968). Political order in changing societies. Yale University Press.

Jong-A-Pin, R. (2009). On the measurement of political instability and its impact on

economic growth. European Journal of Political Economy, 25(1), 15-29.

Londregan, J. B., & Poole, K. T. (1990). Poverty, the coup trap, and the seizure of executive

power. World politics, 42(2), 151-183.

Marshall, M. (2013). Major Episodes of Political Violence (MEPV) and Conflict Regions,

1946-2012. [e-book] Center for Systemic Peace. Available through: Integrated

Network for Societal Conflict Research

http://www.systemicpeace.org/inscr/MEPVcodebook2012.pdf [Accessed: 20 Nov

2013].

Mauro, P. (1995). Corruption and growth. The quarterly journal of economics,110(3), 681-

712.

Murphy, K. M., Shleifer, A., & Vishny, R. W. (1991). The allocation of talent: implications

for growth. The quarterly journal of economics, 106(2), 503-530.

Olson, M. (1982). The rise and decline of nations: Economic growth, stagflation, and social

rigidities. Yale University Press.

Ozler, S., & Tabellini, G. (1991). External debt and political instability (No. w3772).

National Bureau of Economic Research.

The Secretariat of the Cabinet. (2013). Thai Prime Ministers. Available:

http://www.cabinet.thaigov.go.th/bb2_main21.htm. Last accessed 18th Sept 2013.

Shleifer, A., & Vishny, R. W. (1993). Corruption. The Quarterly Journal of

Economics, 108(3), 599-617.

The World Bank. (2013). Indicators. The World Bank Group. http://data.worldbank.org/indicator

[Accessed: 31 Oct 2013].

26

Appendix

Appendix A – Socio-Political Instability (SPI) Index

YEAR GCHANGE COUP ACTOTAL* PSI

1976 1 1 2 1.835

1977 1 1 3 1.002

1978 0 0 3 -0.399

1979 0 0 3 -0.399

1980 1 0 3 1.002

1981 0 0 3 -0.399

1982 0 0 3 -0.399

1983 0 0 3 -0.399

1984 0 0 1 -0.133

1985 0 0 1 -0.133

1986 0 0 1 -0.133

1987 0 0 1 -0.133

1988 1 0 0 1.401

1989 0 0 0 0

1990 0 0 0 0

1991 1 1 0 1.401

1992 1 0 0 2.801

1993 0 0 0 0

1994 0 0 0 0

1995 1 0 0 1.401

1996 1 0 0 1.401

1997 1 0 0 1.401

1998 0 0 0 0

1999 0 0 0 0

2000 0 0 0 0

2001 1 0 0 1.401

2002 0 0 0 0

2003 0 0 1 -0.133

2004 0 0 1 -0.133

2005 0 0 1 -0.133

2006 1 1 1 1.268

2007 0 0 1 -0.133

2008 1 0 1 2.668

2009 0 0 1 -0.133

2010 0 0 1 -0.133

*ACTOTAL = sum of magnitude scores of episodes of international violence, international warfare,

civil violence, civil warfare, ethnic violence, and ethnic warfare

For further information, refer to the Codebook for Major Episodes of Political Violence 2012

(Marshall, Center for Systemic Peace, 2013)

27

Appendix B – Regional Growth (RGROWTH)

The variable RGROWTH was constructed by taking the weighted average of the GDP

growths of ASEAN countries. The weights were based on each country’s GDP (in current

USD) as a proportion of the total.

Of the 10 ASEAN countries, only 5 were used in the calculation of regional growth.

Myanmar was excluded due to the unavailability of data on GDP and GDP growth.

Cambodia and Laos were excluded due to missing variables over the time period and their

relatively small GDPs, 0.61% and 0.39% respectively based on 2010 GDP values. Vietnam

was excluded due to missing variables, and Brunei was excluded due to its relatively small

GDP (0.68% of the total).

2010 GDP (Current USD) Weight

Indonesia 709,266,023,255 0.4192

Malaysia 246,822,585,372 0.1459

Philippines 199,589,448,016 0.1180

Singapore 217,200,123,752 0.1284

Thailand 318,907,930,076 0.1885

Total 1,691,786,110,471 1